基于BRISK特征的动态背景下运动目标检测

针对动态背景下运动目标检测过程中对检测算法实时性和鲁棒性的要求,提出了一种基于二进制鲁棒不变尺度特征(BRISK)的运动目标检测算法.通过改进的BRISK算法检测特征点;为了保证匹配精度和速度,采用K最近邻(KNN)算法进行特征点匹配;运用基于随机抽样一致性(RANSAC)的全局运动参数估计法获取最优全局运动参数;采用帧间差分法进行运动目标检测.实验结果表明:改进的BRISK算法减少了49.8％的特征点数目,KNN算法去除了85.9％的特征点对;在各种场景下能够准确地检测出运动目标,与以往算法相比检测效果较好.

Moving object detection based on BRISK feature in dynamic background

Considering requirements of real-time and robustness of detection algorithm in moving object detection process under dynamic background,a moving object detection algorithm based on binary robust invariant scalable keypoints (BRISK ) is proposed. The improved BRISK algorithm is used to detect feature points. To assure matching precision and speed,the k nearest neighbor(KNN)algorithm is adopted to match feature points. Global motion parameter estimation method based on random sample consensus(RANSAC)is used to obtain the optimal global motion parameters. Interframe difference method is applied to achieve the detection of moving targets. The experimental results indicate that the improved BRISK algorithm reduces the number of feature points by 49. 8％, and KNN algorithm removes 85. 9％ feature points. Meanwhile,the algorithm can accurately detect moving obiects in various scenes,compared with previous algorithm,detecting effect is better.